The impact of a PEER funded research project "Calibration and Verification of OpenSees Models for Simulating the Response through Collapse of Nonplanar RC Walls" is highlighted below. The project Principal Investigator (PI) is Laura N. Lowes, Professor of Civil Engineering, University of Washington, Seattle. The Research Team includes Joshua Stokley, Graduate Student Researcher, University of Washington, Seattle.
Download the Research Project Highlight which includes the abstract (PDF)
Research Impact
The proposed research will provide researchers and practitioners with a calibrated and validated modeling approach for planar and nonplanar concrete walls that enables accurate and computationally efficient simulation of load-deformation response as well as stress, strain, and damage fields using the OpenSees software platform. Currently, a number of approaches exist for use in simulating the nonlinear response of reinforced concrete walls; however, most have significant limitations: i) fiber-type beam-column elements ignore flexure-shear interaction and multiple-vertical-line-element models (MVLEMs) employ simplifying assumptions with respect to this interaction, ii) beam-column elements and MVLEMs require significant assumptions for modeling non-planar walls, iii) solid-element models are typically too computationally intensive for most design and evaluation projects, iv) proprietary software has significant constraints with respect to output variables, load patterns, and solution algorithms, v) few models have been calibrated to provide accurate and mesh-independent simulation of strength loss, and vi) few software include parallelized solution algorithms to support simulations with many degrees-of-freedom. These limitations are not present in the proposed OpenSees modeling approach, which employs layered shell elements (ShellMITC4 and ShellNLDKGQ), the PlaneStressUserMaterial for concrete, and parallelized solution algorithms as needed. Jupyter notebooks and OpenSees scripts will published to enable users to rapidly introduce the modeling approach into their workflows.
Model calibration and sensitivity analysis will be accomplished using the quoFEM software developed by the NHERI SimCenter, which utilizes HPC resources as well as the Dakota software provide by NHERI DesignSafe. Project results will include recommendations for using quoFEM for this type of practical model calibration as well as Jupyter notebooks and analysis scripts to facilitate similar model calibration activities by researchers and practitioners.
Figure 1: Preliminary results of a calibration study for a typical planar wall; note that in (b) experimental data represent the envelope to the measured-load displacement history and simulation results are from a monotonic displacement history.